NCT06973915

Be Right! Back: An Artificial Intelligence Enabled Mobile Application for Patients With Low Back Pain

• Status: Not Yet Recruiting
• Phase: —
• Study type: Observational
• Enrollment: 120 (estimated)
• Sponsor: Singapore General Hospital · Academic / Other
• Sex: All
• Age: 21 Years – 75 Years
• Healthy volunteers: Not accepted

Summary

Low back pain (LBP) is a common problem with complex causes, of which some are modifiable. Physical factors like strength, movement, and pain play a big role, but measuring all these factors accurately is tricky. This is where Artificial Intelligence (AI) comes in. This projects aims to develop an AI solution (in the form of a mobile application) that can measure four key components of the physical factor of LBP, such as how quickly you can stand up five times, your spine's flexibility, how you walk, and your pain levels while moving. The measurements taken by the mobile application will be compared against those of trained physiotherapists to ensure its accuracy. If successful, this AI solution will be a game-changer. Physiotherapists will be able to remotely track the progress of their LBP patients. The data gained from the remote tracking will allow physiotherapists to have a better understanding of the individual profile of each LBP patient and adjust their treatment accordingly, hence allowing for better care and more effective LBP management. In short, this project aims to harness the power of AI to make managing LBP easier for both patients and physiotherapists.

Detailed description

Background: Low back pain (LBP) is a complex condition and its causes are multifactorial, of which the physical, lifestyle, cognitive and emotional factors are potentially modifiable. Due to the complexity of LBP, Artificial Intelligence (AI) can be used to accurately measure and analyze large amounts of data from different sources to aid in the assessment and management of LBP. Objective: Development of an AI model that accurately assesses and measures 4 core components that comprise the Physical factor of LBP. The 4 core components are functional activity (measured using the 5 times sit-to-stand task - 5xSTS), trunk range of motion (ROM), gait pattern and pain levels during movement. Methods: The project aims to recruit 120 LBP patients receiving care at SGH Physiotherapy. For the first (primary) study (n=103), we will compare the measurements (5xSTS, trunk ROM, gait pattern and pain levels during movement) taken by the AI model against that of a trained assessor/physiotherapist. For the second study (n=17), following integration of the AI model with our industry partner's platform, a pilot study will be conducted to assess the feasibility and usability of a minimum viable product. Planned Analysis: For the first study, the Bland-Altman plot will be used to compare the measurements taken by the AI model against that of a trained assessor/physiotherapist. If our hypothesis is correct, the results should show narrow limits of agreement between the 2 methods of measurement. Descriptive statistics will be used for the second study. We anticipate that there will be positive feedback and satisfaction from use of the minimum viable product. Discussion: Successful development of our solution will allow accurate remote tracking of the progress made by LBP patients. This will support/assist physiotherapists in clinical decision-making, hence allowing for more effective management of LBP.

Conditions

• Low Back Pain

Interventions

Type	Name	Description
OTHER	AI model for movement and pain assessment in low back pain	This intervention involves developing an artificial intelligence (AI) model to objectively assess four physical parameters relevant to low back pain (LBP): 1) sit-to-stand performance, 2) trunk range of motion, 3) gait pattern, and 4) facial expression-based pain levels during movement. The AI model processes video recordings of participants performing these tasks to extract movement and facial data, providing standardized measurements. The tool is designed to assist physiotherapists in clinical decision-making by offering consistent and accurate assessments compared to traditional observational methods.

Timeline

• Start date: 2025-06-01
• Primary completion: 2026-09-30
• Completion: 2027-03-31
• First posted: 2025-05-15
• Last updated: 2025-05-15

Locations

1 site across 1 country: Singapore